let me just clarify.
your #1 proposal would effectively mean that for the leaf taxon algorithm, you want “leaves” to be determined based on whatever the most granular taxa are, not to be limited by species as the most granular leaf. moreover, that algorithm would function in the new way, regardless of whether or not the result set was created with the unobserved_by_user_id parameter.
so for example, if this was implemented, and i queried for all mallards at Arthur Storey Park (https://www.inaturalist.org/observations?place_id=164230&subview=grid&taxon_id=6930&view=species), instead of seeing the species Anas platyrhynchos in the species tab, i would see A. p. domesticus in the species tab, since both wild and domestic ducks have been seen at the park, but wild is recorded at species and domestic is recorded at variety, and variety is the lower rank.
so understanding that that’s a potential problematic consequence of your proposal, are you still proposing that as your change? if so, i think you should write up a feature request for this (instead of a bug report) so that the pros and cons can be discussed with the wider community.
if knowing about the problematic consequences of that first proposal has made you decide to abandon that proposal, then let’s talk about your second proposal. because the Species tab in Explore screen is always going to list species that are returned in the Observations tab, if you want Ambrosia trifida to not show up in the Species tab, then that would necessarily mean that A. t. texana would not show in the Observations tab either. in other words, your second proposal would mean that the unobserved_by_user_id would bring back any taxa that you had not observed, unless that taxon was an unobserved subspecies/variety/etc. for which you had observed the species.
is that really what you mean for your second proposal? if so, i would again reframe this as a feature request (rather than a bug report) so that the broader community can chime in on whether this is a desired change or not.
if you’re still thinking about something different, please clarify. or if you’re thinking about a new approach, please share.